search for: salsa20

...s such as PollyDependence, PollyOptimization and PollyCodegen. Especially, I notice that the PollyDependence can lead to significant extra compile-time overhead. Its compile-time percentage for some expensive benchmarks can be summarized as: nestedloop: 41.4% (Polly - Calculate dependence) salsa20: 98.5% (Polly - Calculate dependence) seidel-2d: 72.1% (Polly - Calculate dependence) multiplies: 54.3% (Poly - Calculate dependence) Puzzle: 22.8% (Poly - Calculate dependence) As a result, it is critical to improve the PollyDependence pass. I have previously...

...llyDependence, PollyOptimization and PollyCodegen. Especially, I notice that the PollyDependence can lead to significant extra compile-time overhead. Its compile-time percentage for some expensive benchmarks can be summarized as: > nestedloop: 41.4% (Polly - Calculate dependence) > salsa20: 98.5% (Polly - Calculate dependence) > seidel-2d: 72.1% (Polly - Calculate dependence) > multiplies: 54.3% (Poly - Calculate dependence) > Puzzle: 22.8% (Poly - Calculate dependence) > > > As a result, it is critical to improve the PollyDepe...

...dence, PollyOptimization and PollyCodegen. Especially, I notice that the PollyDependence can lead to significant extra compile-time overhead. Its compile-time percentage for some expensive benchmarks can be summarized as: >>      nestedloop:  41.4% (Polly - Calculate dependence) >>      salsa20:       98.5% (Polly - Calculate dependence) >>      seidel-2d:     72.1% (Polly - Calculate dependence) >>      multiplies:     54.3% (Poly -  Calculate dependence) >>      Puzzle:         22.8% (Poly -  Calculate dependence) >> >> >> As a result, it is critical...

...8 benchmarks improved and 16 benchmarks regressed. Especially, with those recent performance-oriented patch files for ScopDetect/ScopInfo/ScopDependences/..., we have significantly reduced the compile-time overhead of Polly for a large number of benchmarks, such as:     SingleSource/Benchmarks/Misc/salsa20        -97.84%     SingleSource/Benchmarks/Polybench/linear-algebra/solvers/lu/lu        -85.01%     MultiSource/Applications/obsequi/Obsequi        -57.12%     SingleSource/Benchmarks/Polybench/stencils/seidel-2d/seidel-2d        -50.00%     MultiSource/Benchmarks/MiBench/telecomm-gsm/telecomm-gs...

...hmarks improved and 16 benchmarks regressed. Especially, with those recent performance-oriented patch files for ScopDetect/ScopInfo/ScopDependences/..., we have significantly reduced the compile-time overhead of Polly for a large number of benchmarks, such as: > SingleSource/Benchmarks/Misc/salsa20 -97.84% > SingleSource/Benchmarks/Polybench/linear-algebra/solvers/lu/lu -85.01% > MultiSource/Applications/obsequi/Obsequi -57.12% > SingleSource/Benchmarks/Polybench/stencils/seidel-2d/seidel-2d -50.00% > MultiSource/Benchmarks/MiBench/...

...66.6% slowdown SingleSource/Benchmarks/Misc/flops-8 - 64.2% slowdown (from these, I've excluded tests that took less that 0.1 seconds to run). Largest three compile-time slowdowns: MultiSource/Benchmarks/MiBench/security-rijndael/security-rijndael - 1276% slowdown SingleSource/Benchmarks/Misc/salsa20 - 1000% slowdown MultiSource/Benchmarks/Trimaran/enc-3des/enc-3des - 508% slowdown Not everything slows down, MultiSource/Benchmarks/Prolangs-C ++/city/city, for example, compiles 10% faster with vectorization enabled; but, for the most part, things certainly take longer to compile with vectorizat...

...s the first without the patch applied the second with. If this is the case, why does the page say "*** WARNING ***: comparison is against a different machine (pollyScopInfo__clang_DEV__x86_64,11)"? > Results show that this patch file is very effective for several benchmarks, such as salsa20 (reduced by 97.72%), Obsequi (54.62%), seidel-2d (48.64%), telecomm-gsm (33.71%). > >> I did not yet look at the nestedloop benchmark, but it sounds basically >> like a benchmark only consisting of loop nests that we can optimise. >> This is definitely interesting to look into....

...Misc/flops-8 - 64.2% slowdown > Interesting. Do you understand what causes these slowdowns? Can your heuristic be improved? > Largest three compile-time slowdowns: > MultiSource/Benchmarks/MiBench/security-rijndael/security-rijndael - > 1276% slowdown > SingleSource/Benchmarks/Misc/salsa20 - 1000% slowdown > MultiSource/Benchmarks/Trimaran/enc-3des/enc-3des - 508% slowdown Yes, that is a lot. Do you understand if this time is invested well (does it give significant speedups)? If I understood correctly it seems your vectorizer has quadratic complexity which may cause large slow...

...ks improved and 16 benchmarks regressed. Especially, with those recent performance-oriented patch files for ScopDetect/ScopInfo/ScopDependences/..., we have significantly reduced the compile-time overhead of Polly for a large number of benchmarks, such as: >>      SingleSource/Benchmarks/Misc/salsa20        -97.84% >>      SingleSource/Benchmarks/Polybench/linear-algebra/solvers/lu/lu        -85.01% >>      MultiSource/Applications/obsequi/Obsequi        -57.12% >>      SingleSource/Benchmarks/Polybench/stencils/seidel-2d/seidel-2d        -50.00% >>      MultiSource/Ben...

Hello, Full disclosure: my name is Farid Fadaie and I'm in charge of BitTorrent Bleep (a private P2P SIP-based messaging application in early alpha) http://blog.bittorrent.com/2014/07/30/building-an-engine-for-decentralized-communications/ I have personally been a fan of Asterisk and have been using it for years and now that we have (kind of) released Bleep, I wanted to ask you guys to let

...c4 -11.0% MultiSource/Benchmarks/ASCI_Purple/SMG2000/smg2000 -9.0% MultiSource/Applications/minisat/minisat -8.8% External/SPEC/CINT2006/473.astar/473.astar -8.7% SingleSource/Benchmarks/Misc/ReedSolomon -8.6% SingleSource/Benchmarks/BenchmarkGame/nsieve-bits -8.0% SingleSource/Benchmarks/Misc/salsa20 -7.7% MultiSource/Benchmarks/ASC_Sequoia/CrystalMk/CrystalMk -7.4% External/SPEC/CINT2006/401.bzip2/401.bzip2 -7.4% SingleSource/Benchmarks/Misc/mandel-2 -7.3% SingleSource/Benchmarks/Shootout-C++/methcall -6.5% MultiSource/Benchmarks/MiBench/consumer-lame/consumer-lame -6.4% External/SPEC/CF...

...ive vector operations creates inefficient code, and I'll also look at these cases in more detail. > > > Largest three compile-time slowdowns: > > MultiSource/Benchmarks/MiBench/security-rijndael/security-rijndael - > > 1276% slowdown > > SingleSource/Benchmarks/Misc/salsa20 - 1000% slowdown > > MultiSource/Benchmarks/Trimaran/enc-3des/enc-3des - 508% slowdown > > Yes, that is a lot. Do you understand if this time is invested well > (does it give significant speedups)? No, not always. Actually, the security-rijndael test not only takes the longest to...

...e page say "*** WARNING ***: comparison is against a  >different machine (pollyScopInfo__clang_DEV__x86_64,11)"? LNT always report warning if you compare two runs with different tester names. >> Results show that this patch file is very effective for several benchmarks, such as salsa20 (reduced by 97.72%), Obsequi (54.62%), seidel-2d (48.64%), telecomm-gsm (33.71%). >> >>> I did not yet look at the nestedloop benchmark, but it sounds basically >>> like a benchmark only consisting of loop nests that we can optimise. >>> This is definitely interesti...

Hi all, I've started looking at the GlobalMerge pass, enabled by default on ARM and AArch64. I think we should reconsider that, at least for AArch64. As is, the pass just merges all globals together, in groups of 4KB (AArch64, 128B on ARM). At the time it was enabled, the general thinking was "it's almost free, it doesn't affect performance much, we might as well use it".

...rijndael - 114% slowdown MultiSource/Benchmarks/MiBench/network-patricia/network-patricia - 81.8% slowdown SingleSource/Benchmarks/Misc/flops-8 - 79% slowdown Top compile-time slowdowns: MultiSource/Benchmarks/MiBench/security-rijndael/security-rijndael - 832% slowdown SingleSource/Benchmarks/Misc/salsa20 - 600% slowdown MultiSource/Benchmarks/Trimaran/enc-3des/enc-3des - 263% slowdown For this comparison, however (unlike comparing to plain -O3), there are a significant number of compile-time speedups (I guess that this is because vectorization can reduce the number of instructions processed by lat...

...eSource/Benchmarks/Misc/lowercase | * | * | SingleSource/Benchmarks/Misc/flops-8 | * | * | SingleSource/Benchmarks/Misc/ffbench | * | * | SingleSource/Benchmarks/Misc/salsa20 | * | * | SingleSource/Benchmarks/Misc/flops-7 | * | * | SingleSource/Benchmarks/Misc/mandel | * | * | SingleSource/Benchmarks/Misc/perlin...

On 11/08/2011 03:36 PM, Hal Finkel wrote: > On Tue, 2011-11-08 at 12:12 +0100, Tobias Grosser wrote: >> On 11/08/2011 11:45 AM, Hal Finkel wrote: >>> I've attached the latest version of my autovectorization patch. >>> >>> Working through the test suite has proved to be a productive >>> experience ;) -- And almost all of the bugs that it revealed

On Sat, 2011-10-29 at 14:02 -0500, Hal Finkel wrote: > On Sat, 2011-10-29 at 12:30 -0500, Hal Finkel wrote: > > Ralf, et al., > > > > Attached is the latest version of my autovectorization patch. llvmdev > > has been CC'd (as had been suggested to me); this e-mail contains > > additional benchmark results. > > > > First, these are preliminary

...3886|0.158737498| -0.03| 2155883879| 0|0.158739646| -0.03| 2155883879| 0| |SingleSource/Benchmarks/Misc/richards_benchmark.test | 78|0.434660464| 4033193605|0.434746459| -0.02| 4033193598| 0|0.434897056| -0.05| 4033193598| 0| |SingleSource/Benchmarks/Misc/salsa20.test | 40|3.685973193|45487828644|3.687791537| -0.05|45487828691| 0|3.685686752| 0.01|45487828690| 0| |SingleSource/Benchmarks/Misc/whetstone.test | 46|0.784256179| 4409167761|0.784001984| 0.03| 4409167754| 0|...

...  0.84 | 0.97    1.01    n/a          n/a > SingleSource/Benchmarks/Misc/richards_benchmark                          | > 0.0300 5756     0.0100      *                0.0300      |    1.13    1.05 >  1.17 *           1.33 | 1.08    0.97    n/a          n/a > SingleSource/Benchmarks/Misc/salsa20                                     | > 0.0100 2692     0.0100      *                0.0100      |    9.30    9.72 >  9.27 *           9.33 | 0.96    1.00    n/a          n/a > SingleSource/Benchmarks/Misc/whetstone                                   | > 0.0199 3344     0.0199      *    ...